My research is in visual neuroscience. I’m interested in how we recognise things visually. Surely we must know all that already?! No, not by a long way.

We know a lot about the eye and we do a reasonable job of making devices (digital cameras) that do a roughly similar job. But imagine a piece of computer software that took your digital photographs and identified what was happening in them. “This is a photo of your father, sitting on a wooden chair with a carving of a rose on it. Next to him is a Yorkshire terrier chewing a bone.”

For you it’s trivial because of your visual neurons. But the world’s best programmers are nowhere near being able to do that with computers. Google and Facebook do now have face recognition capabilities, and they work reasonably, just as long as the person doesn’t turn their head too far or put on glasses. Would that ever confuse you when you recognise a face? Have you noticed the number of websites that ask you to type in the words that are portrayed in a swirly image? That’s because your recognition is so much better than any automated computer detection system.

We need to know how the brain does such a remarkable job of perceiving the world. That is the endeavour of visual neuroscience.

Within the field I’m particularly interested in the following topics, with further information below.

  • Mid-level vision
  • Face perception
  • Experimental design and software

Mid level Vision

A great deal is known about the initial steps of visual processing. We know that humans have neural mechanisms selectively tuned to simple patterns of particular spatial frequencies and orientations. Much later in the visual pathway, in inferotemporal (IT) cortex, cells respond to extremely complex visual patterns such as images of faces. Very little is known about intermediate levels of visual processing, where early visual signals are presumably combined to represent increasingly complex visual features.

Characterising those intermediate mechanisms is the primary interest of my lab. That work has led us to present psychophysical evidence for visual mechanisms detecting conjunctions, such as detectors for plaids and detectors for curvature.

For review of the work I’ve done in the area, and why I care so much about it, see my review paper, Understanding mid-level representations in visual processing (Peirce, 2015)